Psychology
2012. Vol.3, No.7, 548-554
Published Online July 2012 in SciRes (http://www.SciRP.org/journal/psych) http://dx.doi.org/10.4236/psych.2012.37081
548
Bottle Shape Elicits Gender-Specific Emotion: A Startle Reflex
Modulation Study
Arvina Grahl1, Ursula Greiner1, Peter Walla1,2,3*
1Biological Psychology, Faculty of Psychology, University of Vienna, Vienna, Austria
2School of Psychology, Faculty of Science and Information Technology, University of Newcastle,
Callaghan, Australia
3Applied Neuroscience, Neuroconsult, Vienna, Austria
Received March 1st, 2012; revised April 6th, 2012; accepted May 11th, 2012
Abstract: Does a bottle elicit a gender-specific emotion response that varies as a function of its shape?
The answer to this question is of basic interest as well as of specific interest to the industry. We intro-
duced startle reflex modulation to objectively measure emotion-related aspects of different bottles.
This approach was chosen to complete behavioural data with objective physiological data. Three bot-
tles differing only in shape were used as test stimuli. It is well known that the magnitude of an eye
blink as a response to acoustic startle stimulation reflects current emotion elicited by a foreground
stimulus. Reduced eye blink amplitudes reflect positive emotion, whereas enhanced eye blink ampli-
tudes reflect negative emotion. No significant main effect of bottle shape on eye blink amplitudes was
found, but a significant gender effect occurred for one specific bottle shape. Presentation of a me-
dium-sized, polygonal bottle was associated with significantly larger eye blink amplitudes in male par-
ticipants compared to female participants. We interpret that due to the nature of startle reflex modula-
tion this particular bottle shape elicited a significantly more negative emotion in males than in females.
Interestingly, although both genders rated this bottle least attractive only males demonstrated increased
negative emotion as assessed with startle reflex modulation. It is difficult to further interpret this find-
ing at this stage, but we discuss it in terms of providing evidence that startle reflex modulation is sen-
sitive to subtle emotion-related differences of different bottle shapes. Thus, this method may become
an important tool for product evaluation procedures. We can only speculate on what the gender effect
means in terms of psychology. A link between emotion and the appreciation of aesthetics can be
made.
Keywords: Unbiased Emotion; Bottle Shape; Gender Difference; Startle Reflex Modulation
Introduction
The role of emotion in the appreciation of aesthetics is un-
clear, which seems mainly due to neglecting objective (unbi-
ased) measures of emotion in this context. Usually, subjective
preference is recorded, which can be biased and thus distort
pure emotion-related effects. Our study aimed at demonstrating
that objective and unbiased measures can be taken to show
gender-specific emotion-related differences related to different
bottle shapes. In addition, we aimed at finding a discrepancy
between self report and objective measures like it has been
found in other previous studies (see later).
Imagine a man buys a bottle of wine for dinner. His choice
falls on the one, which in his opinion has the most attractive
design since he has no knowledge about wines whatsoever. At
home his partner is not only disappointed about the wine, but
also asks him if he couldn’t have bought a bottle with at least a
pretty design. Does this ring a bell? This anecdote highlights
well-known variable aspects of perception (and/or the apprecia-
tion of aesthetics), which, inter alia, influence human prefer-
ence (Holbrook & Schindler, 1994). It seems obvious that even
the design of a bottle can be evaluated in different ways de-
pending on various factors including the perceivers’ gender.
For the present study we translated this everyday experience
into an empirical investigation and focussed on gender-specific
objective measures of emotion elicited by differently shaped
bottles. In addition, seeing things from a product developers’
perspective, we thought it could be useful for marketing and
brand management to gain more information about objective
design preferences in comparison to usual subjective prefer-
ences, which can be biased.
Cognitive aspects play a big role regarding preferences, but
in the frame of the present study we focus on emotion-related
aspects. While emotion has certainly always been an important
topic to investigate, also with respect to product design, it has
been very frequently neglected that basic aspects of emotion
may not be conscious and more importantly, they may not be
accessible to consciousness while still influencing behaviour
(see Walla et al., 2002; Walla et al., 2003; Winkielman & Ber-
ridge, 2004; Walla, 2008). Explicitly stated answers to ques-
tions about subjective preference may not result in reliable
information regarding basic rather non-conscious emotion as-
pects. Although we still lack reliable verification that non-
conscious aspects of emotion are better predicting human emo-
tion-based behaviour, there is reason to believe that this is so to
a certain extent and under certain circumstances.
*Corresponding author. Regarding gender differences emotion research revealed

A. GRAHL ET AL.
various significant findings (Biele & Grabowska, 2006; Wa-
ger et al., 2003). Besides behavioural differences related to
emotion perception also different brain activity patterns were
found between males and females (Cahill et al., 2001; Cahill et
al., 2004; Killgore et al., 2001). Thus, there is reason to expect
a gender-specific effect even when emotion elicited by differ-
ently shaped bottles is the question of interest. To the best of
our knowledge, this question has never been answered yet,
especially not by using an objective approach.
In previous studies, qualitative data such as asking for pref-
erences via questionnaires or verbal associations via keyword
listings have been used to quantify human emotion or subjec-
tive preferences. So far, only a few reports are available de-
monstrating the usefulness of objective measures during the
perception of or during the interaction with actual products
(Laparra-Hernandez et al., 2009; Jenkins et al., 2009; Walla et
al., 2010; Geiser & Walla, 2011).
In particular, previous findings (Bradley & Lang, 2000) have
provided evidence of objective ways to measure emotion by
using startle reflex modulation and more recent evidence has
been published that self reported emotion does not always
match objective measures of emotion (Walla et al., 2010;
Geiser & Walla, 2011). In other words, biased emotion va-
lence can be different from unbiased emotion valence even as
responses to the same stimulus. It is this discrepancy together
with the motivation to study gender differences that is of spe-
cial interest here. To collect subjective data (biased emotion
aspects) we planned to record self reported preference data and
among several available objective methods to measure unbiased
emotion aspects we decided to use startle reflex modulation. It
offers insight into aspects of emotion that are not contaminated
by factors such as arousal and/or cognitive interference and
thus provide an excellent new approach to evaluate emotion
aspects related to commercial products.
Startle Reflex Modulation
The first startle reflex research was conducted with pistol
shots and high-speed motion pictures Landis and Hunt, 1939)
in animal populations, mostly rodents. Subsequent research
revealed that the magnitude of the human startle reflex varies as
a function of emotion valence (Bradley et al., 1990; Vrana et al.,
1988). Since this discovery, a large number of papers have been
written reporting about related findings. In principle, the startle
reflex is a defensive reflex to sudden aversive noises. It is
meant to be an automatic protection mechanism of the organ-
ism against potential danger, comparable to the behaviour of
avoidance (escape). Most strikingly, an enhanced startle re-
sponse during any simultaneous foreground stimulation reflects
a negative emotion being elicited by the foreground stimulus,
compared to a reduced startle response which reflects positive
emotion during foreground stimulation. To date, numerous
investigations have been conducted using startle reflex modula-
tion to define emotion aspects; however, this method has been
largely neglected in applied settings, such as product evaluation
and marketing. Only recently, Walla et al. (2011) showed that it
can be used to define emotion aspects related to brand attitude
without demanding any explicit responses from study partici-
pants (that are usually cognitively polluted; biased). This is
particularly interesting with regards to an important methodo-
logical issue. It was initially found that foreground stimuli need
to be above a certain arousal threshold (Cuthbert et al., 1996)
for startle reflex modulation to be sensitive to changes in va-
lence related to them. Although brand attitude as in reading
visually presented brand names is associated with rather low
levels of arousal, startle reflex modulation was obviously sensi-
tive enough to detect significant valence-related differences
reflecting like and dislike related to brands. In addition, va-
lence-related differences were also found during food intake
(Walla et al., 2010), during walking through different virtual
urban environments (Geiser & Walla, 2011), and during driv-
ing through virtual tunnels (Muehlberger et al., 2008) and dur-
ing viewing gradually increasing anger in facial expressions
(Dunning et al., 2010) by using startle reflex modulation. Thus,
there is reason to believe that varying emotion aspects of fore-
ground stimuli such as different bottle shapes can be detected
via this method as well. However, to increase the chance to
detect significant valence-related differences between differ-
ent bottle shapes we used emotion images (true A4 colour
prints) displayed in the background in order to increase the
level of arousal. Participants were told that emotion images
are displayed in the background while they hold each bottle in
their hands to inspect and experience it. No attention had to be
paid to the emotion images during this session of the experi-
ment.
However, in a separate experiment, the emotion images (one
positive and other negative) were also used without simultane-
ous bottle exposure to test whether our mobile recording device
reliably detects valence-related differences through startle re-
flex modulation in a usual setting.
After all, the main goal of the present study was to take a
closer look at emotion aspects related to the perception of dif-
ferent bottles and to detect gender-related differences. To test
our hypothesis, we chose bottles with varying shapes, but same
colour and material. To allow for a comparison between males
and females, we ensured that the study group included equal
numbers of male and female participants.
Method
Participants
Data from 16 university students from different fields were
taken to test the hypothesis that different bottle shapes elicit
different emotion responses (8 female, 8 male). They were
between the age of 20 and 30 years (mean age = 24.88; SD =
2.8). All were right-handed and reported to have no history of
neurological or psychological disorder.
Stimuli
Main foreground stimulation comprised 3 different bottles
made of transparent glass (see Figure 1). These differed only in
shape and size. One bottle was short and pyramid-shaped with
4 edges and a normal neck, another bottle was slightly taller
and with 4 parallel edges and a normal neck and the third was
an even taller, pyramid-shaped bottle with 3 edges and a normal
neck. The participants got one of these bottles at a time and
were instructed to touch and interact with it for 5 minutes.
In addition, we used two images, one emotionally negative
(accident victim) and the other emotionally positive (smiling
woman), to test and confirm the reliability of our method with
respect to its sensitivity related to different emotion valences
(both compared to a neutral condition where only a fixation
cross was presented). These images were color prints (A4)

A. GRAHL ET AL.
Figure 1.
Bottles: Short pyramid-shaped (1), tall with parallel edges (2) and
tall pyramid-shaped (3).
placed in front of all study participants. This was done inde-
pendent from any bottle presentation.
Furthermore, we used these images to increase the level of
arousal during bottle presentation. This idea is based on the fact
that it has been shown that startle reflex modulation works best
for higher arousing foreground stimuli (Cuthbert et al., 1996).
This was done although it has already been shown that even the
intake of different food (Walla et al., 2010), walking through
different virtual urban environments (Geiser & Walla, 2011)
and reading different brand names (Walla et al., 2011) elicit
valence-related differences that can be registered via startle
reflex modulation.
However, all bottle-related data were analysed across emo-
tion images to focus on effects related to differences in bottle
shape only. As a result of this procedure, each bottle was given
to the study participants three times, each time with a different
background emotion image (pleasant, neutral and unpleasant).
The startle stimulus was a 50ms burst of acoustic white noise
delivered via earphones at a sound pressure level of 105 dB
(Walla et al., 2010). Sound pressure level was measured with a
mobile measuring device (produced by Voltcraft). To achieve
respective loudness, a commercial headphone preamplifier was
employed (Behringer; MicroAMP HA400).
Startle Response Meas ur ement
Physiological measurements were taken with a 10-channel
mobile recording device (Nexus10 from Mind Media BV).
Using bipolar electromyography (EMG), muscle potential
changes of the musculus orbicularis oculi of the left eye of
every study participant were measured and stored on the hard
drive of a laptop computer. We used a dual channel electrode
cable with carbon coating and active shielding technology for
low noise and an additional ground electrode cable attached to
the right cheek. The EMG sampling rate was 2048 per s. A
band pass filter from 20 Hz to 500 Hz was applied during
online recording. Raw EMG data were then recalculated by
using the root mean square (RMS) method (epoch-size = 1/16 s)
to transform EMG signals into amplitudes. The resulting am-
plitudes were then subjected to statistical analysis.
Questionnaires
The participants were also asked to complete a simple ques-
tionnaire related to subjective bottle preference. This was done
to also have their self reported subjective preferences regarding
all three bottles (biased emotion).
Procedure
While seated on a comfortable chair, a study participant
completed assessments for demographics and filled in the
questionnaire concerning their subjective preference ratings
related to the bottles. In the following, all sensors were attached
and physiological signals were visually inspected to check for
acceptable signal-to-noise ratios. The actual experiment con-
sisted of 12 consecutive conditions (3 × emotion image and 3
bottles × 3 emotion images), each 5 minutes in duration. All
conditions were counterbalanced between study participants
(random condition order). During each condition, six startle
probes (1 test stimulus for baseline measurement and 5 target
stimuli) were presented with an inter-stimulus interval of 40 s
to 70 s to minimise startle habituation (Filion et al., 1998).
Data Pre-Processing
Raw EMG signals from the orbicularis oculi were filtered,
rectified and computed into amplitudes by the Root-Mean-
Square-method by the software package BioTrace+ provided by
Mind Media BV. Through visual inspection only single ampli-
tudes clearly occurring between 50 ms and 150 ms post-startle
probe were selected for further analysis. This procedure was
done completely unbiased, which means that during the process
of visual inspection the link between data and one of the 12
conditions was not obvious. If several peaks occurred, if the
EMG signal was distorted by activity with an onset prior to the
startle probe or if a response was non-existent altogether, no
startle amplitude was calculated. This procedure resulted in a
total of less than 1% missing values. The amount of startle re-
flex modulation was calculated by dividing the mean of the five
startle amplitudes elicited during lead stimulus presentation by
the baseline amplitude of the corresponding condition, which
was the result of the first test startle stimulus (for each condi-
tion separately).
Statistical Analysis
All mean relative startle response values were subjected to
statistical analysis of variance (ANOVA) and to paired-samples
t-tests. This procedure was done separately for images as fore-
ground stimuli and for bottles as foreground stimuli. For bottles
as foreground stimuli results are reported independent of back-
ground emotion category. Descriptive statistics were drawn up
with subjective preference ratings related to all three bottles.
Results
Physiologi cal Da ta
Emotion Images
A significant main effect of emotion condition [F(2;28) =
3.393; p = 0.047; η² = 0.184] was found. Most importantly, the
difference between the positive and the negative condition
[F(1;15) = 6.929; p = 0.019; η² = 0.207] was significant. This
difference is due to a higher mean startle response associated
with the negative emotion condition (Figure 2). No gender dif-
ferences were found. The pattern of startle responses to emotion
550

A. GRAHL ET AL.
females males
.889
.838
1.071
1.028
1.10
1.00
.90
.80
.70
Mean relative eye blink amplitude
positive negative positive negative
Figure 2.
Mean startle reflex responses to positive and the negative emotion
images for females and males separately. Note that the distribution of
startle responses across emotion conditions is similar between females
and males.
images looked remarkably similar between females and males.
After all, we can conclude that actual emotion valence is re-
flected in our startle response data. This underlines the principle
capacity of startle response modulation, in particular the reli-
able use of our mobile recording device and justifies the use of
this method for the present purpose to measure emotion valence
elicited by bottles with different shapes.
Bottles
All calculations were done with averaged startle responses
across emotion conditions. No significant main effect of bottle
shape on startle responses was found across all study partici-
pants. However, descriptive statistics of females and males
separately demonstrates an obvious startle response difference
in case of bottle 2 (Figure 3). While a repeated measures
ANOVA, including startle responses to all bottle shapes and
“gender” as a between-subject factor revealed no significant
gender effect, one-way ANOVAs between groups revealed a
significant gender difference [F = 4.848; p = 0.045; η² = 0.257]
for bottle 2. The mean startle response for females was 0.96
(SD = 0.15) whereas for males it was 1.11 (SD = 0.13). For
bottle 1 the ANOVA result is (F = 0.067; p = 0.8) and for bottle
3 the result is (F = 0.107; p = 0.749).
Behavioral Data
Following the instruction to name the bottle of highest sub-
jective preference, 7 females and 7 males (out of 16) chose
bottle 3 (88%). Only 1 female participant chose bottle 1 (6%)
and only 1 male participant chose bottle 2 (6%). No gender
difference could be found (see Table 1). Following the instruc-
tion to name the bottle of least subjective preference, 6 females
and 7 males (out of 16) chose bottle 2 (81.25%). Only 2 fe-
males and 1 male rated bottle 1 as the least preferred (see Table
2). Also here, no gender difference could be found.
Discussion
The functional significance of this study is that it provides
evidence for gender-specific unbiased emotion differences
females males
.988
.959
1.039
1.011
1.15
1.10
1.05
1.00
.95
.90
Mean relative eye blink amplitude
1.011
1.114
1 2 3 1 2 3
Figure 3.
Mean startle reflex responses related to the 3 bottles (1, 2 and 3). Note
that only bottle 2 shows a significant gender difference.
Table 1.
High subjective preferences—most preferred bottles (gender) N = 16.
Female Male Male and Female
n % n % n %
Bottle 1 1 12.5 0 0 1 6
Bottle 2 0 0 1 12.5 1 6
Bottle 3 7 87.5 7 87.5 14 88
Table 2.
Low subjective preferences—least preferred bottles (gender) N = 16.
Female Male Male and Female
n % n % n %
Bottle 1 2 251 12.5 3 18.75
Bottle 2 6 757 87.5 13 81.25
Bottle 3 0 0 0 0 0 0
related to different bottle shapes. This has two aspects. First, we
provide evidence that startle reflex modulation as an objective
method is sensitive to subtle valence differences related to de-
sign issues such as bottle shape. Second, we provide evidence
that emotion elicited during confrontation with distinct bottle
shapes can be subject to gender differences. Our study is rather
explorative, but does have some basic science implications.
It is believed that emotion- and cognition-related information
processing provide our brains with merged outputs that drive
decision making and behaviour (Pessoa, 2008; Gray et al.,
2002). For our study this means that subjective preference (or
the appreciation of aesthetics) related to specific objects is
based on the processing of emotion- and cognition-related as-
pects. Even if we believe a decision is made from a logic and
rational perspective, emotion-related information influences our
choices (Gray et al., 2002; Ochsner & Phelps, 2007). Usual
approaches are not able to detect emotion-related responses in
the absence of cognitive influences. The results of our study
provide evidence that unbiased emotion responses can be dif-

A. GRAHL ET AL.
ferent from cognitively influenced preference ratings and thus
provide new insight into emotion impacts of products. In addi-
tion, it demonstrates that one stimulus can actually generate at
least two different emotions.
Recently, several studies about the design of products focus-
sed on showing different affective responses to various product
shapes (Hsia & Chen, 2006). We tried to concentrate on the
emotional component of information processing concerning
different bottle shapes. As mentioned in the introduction, startle
reflex modulation is known for its capacity to measure emotion
valence, in particular, appetitive and aversive effects of any
foreground stimulation (Vrana et al., 1988). We also wanted to
link subjective preference with startle reflex modulation data,
because it has already been found that subjective preference
does not always match objective emotion measures (Geiser &
Walla, 2011; Walla et al., 2010).
In our study we first tested that startle response modulation is
able to distinguish between negative and positive emotion in
our participant cohort and by using a new mobile recording
device. The results clearly demonstrate that this is so. Not only
that, we could also demonstrate that no differences occurred
between females and males, which shows that no gender dif-
ference exists in terms of general image-driven emotion (no
bottles). Given that no general emotion-related gender differ-
ences were found, the following finding related to different
bottle shapes is rather surprising.
Subjective preference related to the three bottles we used
clearly showed a strong preference for bottle 3 (88%) which
was the slim, tall and triangular pyramid-shaped bottle (see
Figure 1). Both females and males rated bottle 3 as most at-
tractive. Bottle 2 on the other hand was rated least attractive
(81.25%) by both females and males. Crucially, although no
subjective preference differences occurred between females and
males, our objective investigation revealed a significant gender
difference between mean startle responses related to bottle 2. In
particular, significantly enhanced eye blink amplitudes in male
participants demonstrate that they were more negatively af-
fected by bottle 2 than female participants. Neither of other
bottles led to any such differences.
Therefore, our study provides evidence that startle reflex
modulation is sensitive to subtle emotion-related (appetitive
versus aversive) variations during confrontation with differently
shaped bottles. This finding denotes the major output of our
study which was intended to be rather explorative than hy-
pothesis-driven. Despite the explorative nature of this study, we
want to speculate on some possible interpretations of our spe-
cific findings. The shape was the only changed parameter be-
tween the bottles we used for this experiment. Thus, it seems
appropriate to infer that only the specific shape of bottle 2 af-
fected male and female participants in different ways. For some
reason, depending on gender, the specific shape of bottle 2
caused different variations of emotion-related information proc-
essing. Comparing the three bottle shapes, it turns out that bot-
tles 1 and 3 do not resemble any bottle commonly-used for the
storage of liquids. Only bottle 2, with its parallel edges and
quadratic appearance, resembles a true and familiar bottle
rather than just an aesthetic object. Further speculation brings
us to the point that the design of bottle 2 resembles a bottle that
usually contains strong alcohol, such as whiskey. It is nothing
more than speculation, but we do wonder whether this could
explain why men were more negatively affected than women
during exposure to bottle 2. Having that said, we would assume
that experience influences objective measures of emotion as
assessed via startle response modulation. This would be against
the idea that startle responses reflect rather unbiased emotion.
Or it could be the result of subconscious conditioning effects
that can bias basic emotion-related information processing.
However, it is just a speculation anyway. Former studies
showed a gender difference in behaviour related to alcohol
consumption (Kessler et al., 1994). Usually, it is mentioned that
men have a higher alcohol dependence risk than women. If this
evidence is linked with our findings, we may interpret that star-
tle reflex modulation data somehow reflect differences in alco-
hol dependence risk on a subliminal level. Certainly, a fol-
low-up study is needed to test this idea.
From a more design-related perspective, it should be men-
tioned that men were found to be more image-driven in their
preferences for specific designs than women (Xue & Yen,
2007), but with the caveat that further studies will be needed to
strengthen this expectation. Assuming that our speculation re-
garding the link between bottle 2 and alcohol is correct, one
could argue that this image-driven influence resulted in the
lowest subjective preference for bottle 2 in men. In our study,
in men subjective preference and objective measurement
matches. In women, this is not the case. Although none of the
female startle responses differed significantly between the three
bottles in terms of analytic statictics, women had the most ap-
petitive motivation (smallest eye blink response) while experi-
encing bottle 2. However, women also rated bottle 2 as the least
attractive, which is evidence for a discrepancy between subjec-
tive preference and objective measures of emotion, a situation
that already occurred in previous studies (Geiser & Walla, 2011;
Walla et al., 2010).
It is further concluded that our findings show that subjective
preference related to bottle shapes involves other brain proc-
esses than those measured through startle reflex modulation.
However, both measures are claimed to pick up emotion-related
information. Whatever aspect of an emotion more obviously
drives human behaviour, it may be useful to collect both as-
pects if, for example, a customer is asked for his disposition to
buy a specific product. This link gives this study its applied
character. Crilly et al. (2004) created a framework for design as
a process of communication between producers and consumers.
It involves a component called “context of consumption” in
which they differentiate the consumers’ response to a specific
product into cognition, affect and behaviour. Their view of
cognition could be especially interesting for our study. They
separate cognition into aesthetic impression, semantic interpre-
tation and symbolic association. In addition, they differentiate
between objective and subjective information with respect to an
aesthetic impression related to a specific product. Subjectivity
and objectivity related to a specific product is included in their
framework; however an effect of emotion on aesthetic impress-
sion or subjective preference is not integrated. It is mentioned
that subjective information is seen as the consumer’s experi-
ence and their familiarity with the product, including its shape.
This relates to our idea in that men see the second bottle in a
different way than women, because of its familiarity and shape.
By including objective measures of emotion as a fourth com-
ponent of a consumer’s response to products, producers and
designers could gain important information. All of the compo-
nents of their framework also seem to be influenced by differ-
ent characteristics. In this context, Crilly et al. (2004) also men-
tioned gender similar to us.
552

A. GRAHL ET AL.
The relationship between product-design (or shape) and the
emotion caused by it formed a part of some other recent studies,
e.g. (Hekkert & Desmet, 2002; Hekkert & Desmet, 2007; Ho &
Sju, 2010). However, most often, emotion was seen equivalent
to “feelings” (e.g. love or anger) evoked by a product or as
signals concerning the appraisal of favourable or harmful
events. It is suggested that our study points to a new direction
by involving physiological components through measuring emo-
tion via startle reflex modulation. The future may bring us to
the point where we replace the term emotion with affective
informative processing to highlight non-conscious aspects of
emotion-related information processing. Further, we will have
to accept that several emotions can occur simultaneously.
Other methodological approaches in the field of product de-
sign have however already revealed that, depending on gender
for instance, certain differences exist with regard to subjective
preferences. For example, one group (Moss & Colman, 2001)
used modified business and Christmas card designs to show a
significant difference between male and female subjective
preferences. Others (Xue & Yen, 2007) attempted to demon-
strate a gender difference in the perception of designs by using
ten different mobile phones, MP3 players and fragrance bottles.
Their qualitative study involved 72 participants and paired se-
lected design objects with a number of product features related
to aesthetics, function and social value. To evaluate partici-
pants’ choices, a number of schemes, references and associated
expressions for keywords were employed. They found that men
were more concerned with the overall structure of a product,
whereas women paid more attention to organic forms, details
and textures. Both genders were in agreement regarding the
simplicity of a design, although men seemed to view it in more
image-related terms, whereas women favoured it over practical
issues.
A common method is to study preferences via a collection of
keywords and associations. Prior to choosing a product, it is
necessary to think about its function, costs, utility, aesthetics
and so on. However, every choice is also affected by emotion
(as mentioned above). What does “emotion” mean at all? A list
of 92 definitions classified into 11 categories (Kleinginna &
Kleinginna, 1981) nicely demonstrates common disagreement
concerning emotion and its definition. To date, researchers have
still been unable to arrive at a simple common definition.
However, one would think that exploring emotion requires an
idea about what it actually is. Most commonly emotions are
explained to be a complex amalgam of behaviour, cognitions,
physiological changes and feelings. Thus, emotion is inter-
changeably used for different things. We suggest that, under
certain circumstances, affective information processing may
remain entirely non-conscious even when the person is attentive
and motivated to describe his or her feelings correctly (see Ber-
ridge & Winkielman, 2003). Such affective processing should
not be labeled an emotion, although it may drive a person’s
behaviour while remaining inaccessible to conscious aware-
ness. In short, we propose the existence of genuinely non-con-
scious affective information processing. In line with that notion,
we believe that startle response modulation measures is sensi-
tive to non-conscious affective processing.
Acknowledgements
The authors want to thank Cathal Whelehan for reading the
manuscript and providing helpful comments.
REFERENCES
Berridge, K. C., & Winkielman, P. (2003). What is an unconscious
emotion? The case for unconscious “liking”. Cognition and Emotion,
17, 181-211. doi:10.1080/02699930302289
Biele, C., & Grabowska, A. (2006). Sex differences in perception of
emotion intensity in dynamic and static facial expressions. Experi-
mental Brain Research, 171, 1-6. doi:10.1007/s00221-005-0254-0
Bradley, M. M., Cuthbert, B. N., & Lang P. J. (1990). Startle reflex
modification: Emotion or attention? Psychophysiol ogy, 27, 513-522.
doi:10.1111/j.1469-8986.1990.tb01966.x
Bradley, M. M., & Lang, P. J. (2000). Measuring emotions: Behaviour,
feeling and physiology. In: R. Lane, & L. Nadel (Eds.), Cognitive
neuroscience of emotion (pp. 242-276). New York: Oxford Univer-
sity Press.
Cahill, L., Haier, R. J., White, N. S., Fallon, J., Kilpatrick, L., Lawrence,
C., Potkin, S. G., & Alkire, M. T. (2001). Sex-related difference in
amygdala activity during emotionally influenced memory storage.
Neurobiol. Learning and Memory, 75, 1-9.
doi:10.1006/nlme.2000.3999
Cahill, L., Uncapher, M., Kilpatrick, L., Alkire, M. T., & Turner, J.
(2004). Sex-related hemispheric lateralisation of amygdala function
in emotionally influenced memory: An fMRI investigation. Learning
and Memory, 11, 261-266. doi:10.1101/lm.70504
Crilly, N., Moultrie, J., & Clarkson, P. J. (2004). Seeing things: Con-
sumer response to the visual domain in product design. Design Stud-
ies, 25, 547-577. doi:10.1016/j.destud.2004.03.001
Cuthbert, B. N., Bradley, M. M., & Lang, P. J. (1996). Probing picture
perception: Activation and emotion. Psychophysiology, 33, 103-111.
doi:10.1111/j.1469-8986.1996.tb02114.x
Dunning, J., Auriemmo, A., Castille, C., & Hajcak, G. (2010). In the
face of anger: Startle modulation to graded facial expressions. Psy-
chophysiology, 47, 874-878.
Filion, D. L., Dawson, M. E., & Schell, M. (1998). The psychological
significance of human startle eye blink modification: A review. Bio-
logical Psychology, 47, 1-43.
doi:10.1016/S0301-0511(97)00020-3
Geiser, M., & Walla, P. (2011). Objective measures of emotion during
virtual walks through urban environments. Applied Scien ce s , 1, 1-11.
Gray, J. R., Braver, T. S., & Raichle, M. E. (2002). Integration of emo-
tion and cognition in the lateral prefrontal cortex. Proceedings of the
National Academy of Scienc e s USA, 99, 4115-4120.
doi:10.1073/pnas.062381899
Hekkert, P., & Desmet, P. M. A. (2002). The basis of product emotions.
In W. S. Green, & P. W. Jordan (Eds.), Pleasure with products, be-
yond usability (pp. 61-67). London: Taylor & Francis.
doi:10.1201/9780203302279.ch4
Hekkert, P., & Desmet, P. M. A. (2007). Framework of product ex-
perience. International Journal of Design, 1, 57-66.
Ho, A. G., & Siu, K. W. M. (2010). Role of designers in the new per-
spective of design and emotion. Design principles and practices: An
International Journal, 4 , 15-24.
Holbrook, M. B., & Schindler, R. M. (1994). Age, sex, and attitude
toward the past as predictors of consumers’ aesthetic tastes for cul-
tural products. Journal of Mark eti ng Research, 31, 412-422.
doi:10.2307/3152228
Hsiao, K. A., & Chen, L. L. (2006). Fundamental dimensions of affec-
tive responses to product shapes. International Journal of Industrial
Ergonomics, 36, 553-564. doi:10.1016/j.ergon.2005.11.009
Jenkins, S. D., Brown, R. D. H., & Rutterford, N. (2009). Comparing
thermographic, EEG, and subjective measures of affective experi-
ence during simulated product interactions. International Journal of
Design, 3, 53-65.
Kessler, R. C., McGonagle, K. A., & Zhao, S. (1994). Lifetime and 12-
month prevalence of DSM-III-R psychiatric disorders in the United
States. Archives of General Psychiatry, 51, 8-19.
doi:10.1001/archpsyc.1994.03950010008002
Killgore, W. D., Oki, M., & Yurgelun-Todd, D. A. (2001). Sex-specific
developmental changes in amygdala responses to affective faces.
Neuroreport, 12, 427-433.
doi:10.1097/00001756-200102120-00047

A. GRAHL ET AL.
554
Kleinginna, P. R. Jr., & Kleinginna, A. M. (1981). A categorized list of
emotion definitions, with suggestions for a consensual definition.
Motivation and Emotion, 5, 345-379. doi:10.1007/BF00992553
Landis, C., & Hunt, W. A. (1939). The startle pattern. New York:
Farrar.
Laparra-Hernandez, J., Belda-Lois, J. M., Medina, E., Campos, N., &
Poveda, R. (2009). EMG and GSR signals for evaluating user’s per-
ception of different types of ceramic flooring. International Journal
of Industrial Ergonomics, 39, 326-332.
doi:10.1016/j.ergon.2008.02.011
Moss, G., & Colman, A. M. (2001). Choices and preferences: Experi-
ments on gender differences. Brand Management, 9, 89-99.
doi:10.1057/palgrave.bm.2540057
Muehlberger, A., Wieser, M. J., & Pauli, P. (2008). Darkness-enhanced
startle responses in ecologically valid environments: A virtual tunnel
driving experiment. Biological Psychology, 77, 47-52.
doi:10.1016/j.biopsycho.2007.09.004
Ochsner, K. N., & Phelps, E. (2007). Emerging perspectives on emo-
tion-cognition interactions. Trends in Cognitive Sciences, 11, 317-
318. doi:10.1016/j.tics.2007.06.008
Pessoa, L. (2008). On the relationship between emotion and cognition.
Nature Review Neuroscience, 9, 148-158. doi:10.1038/nrn2317
Vrana, S. R., Spence, E. L., & Lang, P. J. (1988). The startle probe
response: A new measure of emotion? Journal of Abnormal Psy-
chology, 97, 487-491. doi:10.1037/0021-843X.97.4.487
Wager, T. D., Phan, K. L., Liberzon, I., & Taylor, S. F. (2003). Valence,
gender, and lateralisation of functional brain anatomy in emotion: A
meta-analysis of findings from neuroimaging. Neuroimage, 19, 513-
531. doi:10.1016/S1053-8119(03)00078-8
Walla, P., Hufnagl, B., Lehrner, J., Mayer, D., Lindinger, G., Deecke,
L., & Lang, W. (2002). Evidence of conscious and subconscious ol-
factory information processing during word encoding: A magne-
toencephalographic (MEG) study. Cognitive Brain Research, 14,
309-316. doi:10.1016/S0926-6410(02)00121-0
Walla, P., Hufnagl, B., Lehrner, J., Mayer, D., Lindinger, G., Imhof, H.,
Deecke, L., & Lang, W. (2003). Olfaction and depth of word proc-
essing: A magnetoencephalographic (MEG) study. Neuroimage, 18,
104-116. doi:10.1006/nimg.2002.1324
Walla, P. (2008). Olfaction and its dynamic influence on word and face
processing: Cross-modal integration. Progress in Neurobiology, 84,
192-209. doi:10.1016/j.pneurobio.2007.10.005
Walla, P., Richter, M., Färber, S., Leodolter, U., & Bauer, H. (2010).
Food-evoked changes in humans. Startle response modulation and
event-related brain potentials (ERPs). Journal of Psychophysiology,
24, 25-32. doi:10.1027/0269-8803/a000003
Walla, P., Brenner, G., & Koller, M. (2011). Objective measures of
emotion related to brand attitude: A new way to quantify emotion-
related aspects relevant to marketing. PL oS ONE, 6, e26782.
doi:10.1371/journal.pone.0026782
Winkielman, P., & Berridge, K. C. (2004). Unconscious emotion. Cur-
rent Directions in Psych ological Science, 13, 120-123.
doi:10.1111/j.0963-7214.2004.00288.x
Xue, L., & Yen, C. C. (2007). Towards female preferences in design—
A pilot study. I nternational Journal of Design, 1, 11-27.